Rating Opponent Sportsmanship?

Sort:
OldMandowntheRoad
Excellent idea!!!! I would love to see that, like a karma rating.
Pimpingpawnage
batgirl wrote: Now we're talking turkey.

You certainly are!!

*recoils from laptop after nasty shock*

mary123

First, I would like to say that it is very hard to improve what is a completely outstanding site/system. Chess.com rocks, and rocks absolutely. So: thank you sooo sooo much for it.

And: I worry about the nanny-state. Yeah, some people (and only some) are rude: a) won't resign lost games (why?!), b) are rude in the msg box c) all friendly till they start to lose, and then fall silent/don't thank you for the game, etc.  My solution is: I don't play them again/won't add them to friends.

The idea is: we are all big people in this big wide world, and we should be grown up enuf to look after ourselves, and just get over the fact that not everyone is warm and cuddly in the way we imagine ourselves to be. 

That said: If you do implement it (and the logarith suggested sounds deeply considered), I'd suggest that no rating be given until someone has played 10 games. There is a lot to get to know on the site, and the rating stuff would be an impost for beginners.

 But: overwhelmingly, thanks to the guys and gels behind this boon. 

 

 

 


JediMaster
I would like to add something.  I am not sure exactly how you would rate chess players, but if you rate them I think it would only be right and fair to not just rate on a numerical system, but to actually type in a few words giving specifics of your reasons of how you came to the way you rated your opponent.
shadowc
JediMaster wrote: I would like to add something.  I am not sure exactly how you would rate chess players, but if you rate them I think it would only be right and fair to not just rate on a numerical system, but to actually type in a few words giving specifics of your reasons of how you came to the way you rated your opponent.

 I'm afraid it would be impossible to automatically track anything that a user wants to say about his oponent, given the obvious complexities of the human language.


justwrightin
I just don't know... Personally, I think that if someone keeps on winning they will have a lower rating than someone who keeps losing. Expecially if potential cheating is involved. Honestly, who in the hell can go 160-0 without a little CPU help. In this case maybe the sportsmanship would be a good thing.
Vance917

There does seem to be a certain irony here, and I am pointing this out not to give you a hard time, but rather to hopefully clarify your point.  By way of analogy, some individuals preach tolerance for all, with the hidden caveat that these individuals do not tolerate those they find to be intolerant.  Likewise, you are opposed to opinionated posters who try to impose their views on others.  Is this not your opinion, and one that you are now trying to impose on others?  The paradox is that if you advocate for complete freedom to do as you will, unconstrained by consideration for others, then you are empowering others to curtail your freedom; they are, by definition, free to do that.  So do we want complete anarchy, or do we want some rules of decency?

tarrasch

I think this is a great idea, and will be very useful.

People have said that there might be a problem because of different opinions and views on what good means. But, really, this system's purpose is not to exactly rate the "goodness" in a person, in only needs to asses whether a player can play a game of chess in decent conditions, without being rude, dragging out completely lost games etc. So, if your opponent doesn't do anything really annoying, that means that they're able to play a game in decent conditions and deserve a good rating.

Perhaps it would be better to rephrase the question to "Did your opponent display an unpleasant behaviour during the game?".

IMO, you should just turn the system on right now, and see what happens. Maybe just for live chess, because things go much faster, and then see what problems arise, if any.

 

Regarding sore losers and jerks, their votes won't be worth much after giving a few bad ratings.

 

Also, the system could also work like reputation on most forums, so that you can only rate players if your get over a certain rating, so only the good guys rate. This would be a lot simpler, and appears to have no drawbacks ( to me, at least ).

Vance917

To be completely valid, one would have to also rate the raters.  As in, do some raters always take a hostile position?  Are others too benevolent?  Do we take these predispositions into account somehow with statistical adjustment?  It's a great idea, but doing it right would be complicated.

tarrasch

" As in, do some raters always take a hostile position? "

Already done, look at the first posts closer.

Vance917

I know that I was not the first one to think of that possibility, but have we come up with a satisfactory solution?

Vance917

There is a body of research on ranking teams, and selecting the best teams, based upon incomplete pairings (that is, not a round robin -- some teams never face each other).  So some teams have tougher schedules than others, and this needs to be considered in the rankings.  Likewise here, not every player will be rated by every rater, so some players will have, overall, tougher sets of raters than others.  I cannot recall the entire solution, but I do recall that it involved the eigenvectors of the matrix of game results.  It was pretty neat (at least to a geek like me), because when I first learned about eigenvectors, I couldn't imagine that they would ever have any practical application.

scotterpop

It would seem that something like a strict percentage would be fine.  For example, if someone has played 100 games, and had 3 "good" and 2 "bad" then they would have a +3%/-2% score.

Regardless of what system is chosen, I'm certain it will be changed over time.  Whatever you choose to start with, put in just enough effort to get it rolling and then be ready to alter it...that's my suggestion.

Any system, in this case, is better than none.

tarrasch
AnthonyCG wrote:
Vance917 wrote:

I know that I was not the first one to think of that possibility, but have we come up with a satisfactory solution?


I doubt it....

What can they really do about it?


The more bad ratings you give, the less they matter. So, if you always give a bad rating, your rating won't matter at all.

95RITZ9BITZ

Great idea! I encorage!

Cystem_Phailure
echecs06 wrote:

Imagine students rating their teachers: 


I remember some of the ratings filled out for me by university students at the end of semesters.  Once in a while there would be one with all negatives, all the way down the line.  Well, that's their opinion, right?  Except we used standardized rating forms, and not all questions were always applicable to all classes, yet they'd still be scored negatively.  Or, for instance, I'd have a class where all semester long I returned every single exam and problem set during the very next class session, and yet a student would give the most negative mark possible for "exams and assignments were returned in a timely manner".  Clearly a case of someone using ratings inapprpriately as revenge for some other problem they have, whether real or imagined.

There will always be people who use negative ratings to settle a score, or avenge some perceived insult, or just lash out because they didn't like their lunch that day.  Requiring a certain "reputation" before someone can bestow a rating is also useless, because cliques form and people increase one another's rep through deals.

I hope chess.com doesn't go the route of ratings or reputation.  It won't work any better here than it has anywhere else.  It's a silly system that doesn't solve anything, and would only appease a small minority who would enjoy the illusion that they have been given some authority to administer justice to anyone they think looks at them the wrong way.

Vance917
scotterpop wrote:

It would seem that something like a strict percentage would be fine.  For example, if someone has played 100 games, and had 3 "good" and 2 "bad" then they would have a +3%/-2% score.

Regardless of what system is chosen, I'm certain it will be changed over time.  Whatever you choose to start with, put in just enough effort to get it rolling and then be ready to alter it...that's my suggestion.

Any system, in this case, is better than none.


 Cystem and Senior, you both raise valid points.  Any system would be subject to manipulation and abuse, and it could backfire.  But at the same time, Scott also has a valid point, in that sometimes perfection is the enemy of the good, and a first step, though not perfect, still represents progress.  Is that the case here?  Well, I don't know.  It could be.  Or not.  But whether it is or not is specific to this case, and cannot be ascertained by general considerations.

acbell1996

Great idea. all for it.

-X-

I say let's try it. If there are wrinkles we'll work them out!

bondocel

I don't know if Erik or some other staff member will see my post. I wanted to add that on another backgammon site (3dfibs), there is a similar system called RepBot. I have no clear idea how it works, but I guess you may want to take a look, it is running for some time.